DAMAS with Compression Computational Grid for Acoustic Source Mapping

Wei Ma,Xun Liu

DOI: https://doi.org/10.1016/j.jsv.2017.03.027

IF: 4.7

2017-01-01

Journal of Sound and Vibration

Abstract:Nowadays phased microphone arrays have become a standard technique for acoustic source mapping. Compared with the conventional delay-and-sum beamforming method, deconvolution approaches such as DAMAS successfully improve the spatial resolution, however require high computational effort. Without optimizing DAMAS algorithm, recently DAMAS with wavelet compression computational grid (denoted by DAMAS-CG1) has reduced significantly computational run time of DAMAS in applications (Ma & Liu, J. Sound Vib., 395, 2017), however DAMAS-CG1 has an inevitable deficiency that the occurrence probability of aliasing increases slightly for complicated sound source. This paper proposes a novel algorithm that DAMAS with computational grid compressed by discarding grid points with non-positive beamforming (denoted by DAMAS-CG2). Application simulations and an airfoil trailing edge noise experiment show that DAMAS-CG2 not only reduces significantly the computational run time but also retains the spatial resolution of DAMAS on original grid. Moreover DAMAS-CG2 overcomes the inevitable deficiency of DAMAS-CG1, and is simpler and more practical than DAMAS-CG1, although DAMAS-CG2 is usually less effective than DAMAS-CG1.